Video satellite imagery is a new technique for earth dynamic observation and has a wide range of uses in environmental fields. Despite its capability of dynamic targets' detection, it sustains a serious restriction of the image quality due to the degradation and compression in its imaging process. Hence, the super-resolution (SR) reconstruction on these compressed low-spatial-resolution images is of significance to afterward ground objects recognition and detection tasks. Based on the recent proposed state-of-the-art convolutional neural networks (CNNs) SR methods, we proposed an SR method which could get more precise reconstructed high-spatial-resolution images. Trained with Gaofen-2 satellite images, a robust CNN model specified in satellite image SR is obtained. Experimentally, the reconstruction results on Jilin-1 mission satellite images validate the effectiveness of our method.

Convolutional neural networks (CNNs) have wide applications in pattern recognition and image processing. Despite recent advances, much remains to be done for CNNs to learn a better representation of image samples. Therefore, constant optimizations should be provided on CNNs. To achieve a good performance on classification, intuitively, samples' interclass separability, or intraclass compactness should be simultaneously maximized. Accordingly, in this paper, we propose a new network, named separability and compactness network (SCNet) to rectify this problem. SCNet minimizes the softmax loss and the distance between features of samples from the same class under a jointly supervised framework, resulting in simultaneous maximization of interclass separability and intraclass compactness of samples. Furthermore, considering the convenience and the efficiency of the cosine similarity in face recognition tasks, we incorporate it into SCNet's distance metric to enable sample features from the same class to line up in the same direction and those from different classes to have a large angle of separation. We apply SCNet to three different tasks: visual classification, face recognition, and image superresolution. Experiments on both public data sets and real-world satellite images validate the effectiveness of our SCNet.

Chlorophyll-a (Chl-a) concentration is considered as a key indicator of the eutrophic status of inland water bodies. Various algorithms have been developed for estimating Chl-a in order to improve the accuracy of predictive models. The objective of this study is to assess the potential of hyperspectral multi-band indices to estimate the Chl-a concentration in Dianshan Lake, which is the largest lake in Shanghai, an international metropolis of China. Based on field spectral measurements and in-situ Chl-a concentration collected on 7–8 September 2010, hyperspectral multi-band indices were calibrated to estimate the Chl-a concentration with optimal wavelengths selected by model tuning. A three-band index accounts for 87.36% (R² = 0.8736) of the Chl-a variation. A four-band index, which adds a wavelength in the near infrared (NIR) region, results in a higher R² (0.8997) by removing the absorption and backscattering effects of suspended solids. To test the applicability of the proposed indices for routinely monitoring of Chl-a in inland lakes, simulated Hyperion and real HJ-1A satellite data were selected to estimate the Chl-a concentration. The results show that the explanatory powers of these satellite hyperspectral multi-band indices are relatively high with R² = 0.8559, 0.8945, 0.7969, and 0.8241 for simulated Hyperion and real HJ-1A satellite data, respectively. All of the results provide strong evidence that hyperspectral multi-band indices are promising and applicable to estimate Chl-a in eutrophic inland lakes.

Federated learning allows for cooperative training among distributed clients by sharing their locally learned model parameters, such as weights or gradients. However, as model size increases, the communication bandwidth required for deployment in wireless networks becomes a bottleneck. To address this, we propose a residual-based federated learning framework (ResFed) that transmits residuals instead of gradients or weights in networks. By predicting model updates at both clients and the server, residuals are calculated as the difference between updated and predicted models and contain more dense information than weights or gradients. We find that the residuals are less sensitive to an increasing compression ratio than other parameters, and hence use lossy compression techniques on residuals to improve communication efficiency for training in federated settings. With the same compression ratio, ResFed outperforms current methods (weight- or gradient-based federated learning) by over $1.4\times $ on federated data sets, including MNIST, FashionMNIST, SVHN, CIFAR-10, CIFAR-100, and FEMNIST, in client-to-server communication, and can also be applied to reduce communication costs for server-to-client communication.